The overall purpose of our research is to establish how different dietary patterns affect antioxidant capacity and free-radical mediated cellular damage. In a clinical trial, human volunteers are randomized to one of three diets that vary in fat content and amounts of fruits and vegetables. Free-radical mediated damage is assessed with measures of lipid peroxidation and is related to the antioxidant capacity of the blood. Our research has been conducted as an ancillary study within the context of an NHLBI-sponsored clinical trial entitled Dietary Approach to Stop Hypertension. The design of the study is a randomized, placebo- controlled clinical trial. Volunteer participants with borderline to mild hypertension, and no other concomitant disease, are eligible for this study. The trial was designed to look at the effects of diets that vary in fat content and the amounts of fruits and vegetables on blood pressure. Patients who are interested and eligible are all placed for three weeks on a diet that is 37% fat and is low in fruits and vegetables. After three weeks, patients were randomized to one of three diets, either the baseline diet, a diet that is low in fat (25%) and high in fruits and vegetables, or on a diet that has 37% fat but is high in fruits and vegetables. The main outcome variable for the trial is blood pressure measurements. The objective of our ancillary study is to determine the effects of diet on measures of the precursors of atherosclerosis, including the collection and analysis of breath ethane and serum to determine the extent of lipid peroxidation. Our hypothesis is that diets high in fruits and vegetables which have an abundance of exogenous antioxidants, are protective against free-radical mediated damage as reflected by peroxidation of lipids.. Outcome variables will also be correlated with baseline atherosclerotic cardiovascular disease risk factors. In June of 1996, we completed the last of our five cohorts. There was a tremendous effort in recruitment, which was highly successful. We enrolled and sampled 123 participants overall, 35 in the last cohort during the most recent fiscal year. These numbers exceed our a priori expectations, on which we based our power calculations. Hence we have completed data collection, and we are currently performing data analysis that to date has resulted in one submitted publication and one in preparation.